Skip to main content

Advertisement

SpringerLink
Log in

Maximum likelihood from spatial random effects models via the stochastic approximation expectation maximization algorithm

  • Published:
Statistics and Computing Aims and scope Submit manuscript

Abstract

We introduce a class of spatial random effects models that have Markov random fields (MRF) as latent processes. Calculating the maximum likelihood estimates of unknown parameters in SREs is extremely difficult, because the normalizing factors of MRFs and additional integrations from unobserved random effects are computationally prohibitive. We propose a stochastic approximation expectation-maximization (SAEM) algorithm to maximize the likelihood functions of spatial random effects models. The SAEM algorithm integrates recent improvements in stochastic approximation algorithms; it also includes components of the Newton-Raphson algorithm and the expectation-maximization (EM) gradient algorithm. The convergence of the SAEM algorithm is guaranteed under some mild conditions. We apply the SAEM algorithm to three examples that are representative of real-world applications: a state space model, a noisy Ising model, and segmenting magnetic resonance images (MRI) of the human brain. The SAEM algorithm gives satisfactory results in finding the maximum likelihood estimate of spatial random effects models in each of these instances.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

References

  • Aitkin M. 1996. A general maximum likelihood analysis of overdispersion in generalized linear models. Statistics and Computing 6: 251–262.

    Article  Google Scholar 

  • Benveniste A., Métivier M., and Priouret P. 1990. Adaptive Algorithms and Stochastic Approximations. Springer-Verlag, New York.

    MATH  Google Scholar 

  • Besag J.E. 1974. Spatial interaction and the statistical analysis of lattice systems (with discussion). Journal of Royal Statistical Society, Series B 36: 192–236.

    MATH  MathSciNet  Google Scholar 

  • Besag J.E. 1986. On the statistical analysis of dirty pictures (with discussion). Journal of Royal Statistical Society, Series B 48: 259–302.

    MATH  MathSciNet  Google Scholar 

  • Besag J.E., Green P., Higdon D., and Mengersen K. 1995. Bayesian computation and stochastic systems (with discussion). Statistical Science 10: 3–66.

    MATH  MathSciNet  Google Scholar 

  • Breslow N.E. and Clayton D.G. 1993. Approximate inference in generalized linear mixed models. Journal of American Statistical Association 88: 9–25.

    Article  MATH  Google Scholar 

  • Chan K.S. and Ledolter J. 1995. Monte Carlo EM estimation for time series models involving counts. Journal of American Statistical Association 90: 242–252.

    Article  MATH  MathSciNet  Google Scholar 

  • Christensen O.F. and Waagepetersen R.P. 2002. Bayesian prediction of spatial count data using generalized linear mixed models. Biometrics 58: 280–286.

    Article  MathSciNet  Google Scholar 

  • Delyon B., Lavielle E., and Moulines E. 1999. Convergence of a stochastic approximation version of the EM algorithm. Annals of Statistics 27: 94–128.

    Article  MATH  MathSciNet  Google Scholar 

  • Derin H. and Elliott H. 1987. Modeling and segmentation of noisy and textured images using Gibbs random fields. IEEE Transactions on Pattern Analysis and Machine Intelligence 9: 39–55.

    Google Scholar 

  • Diggle P.J., Tawn J.A., and Moyeed R.A. 1998. Model-based geostatistics (with discussion). Applied Statistics 47: 299–350.

    MATH  MathSciNet  Google Scholar 

  • Durbin J. and Koopman S.J. 1997. Monte Carlo maximum likelihood estimation for non-Gaussian state space models. Biometrika 84: 669–684.

    Article  MATH  MathSciNet  Google Scholar 

  • Durbin J. and Koopman S.J. 2000. Time series analysis of non-Gaussian observations based on state space models from both classical and Bayesian perspectives (with discussion). Journal of Royal Statistical Society, Series B 62: 3–56.

    Article  MATH  MathSciNet  Google Scholar 

  • Gelman A. and Meng X.L. 1998. Simulating normalizing constants: from importance sampling to bridge sampling to path sampling. Statistical Science 13: 163–185.

    Article  MATH  MathSciNet  Google Scholar 

  • Geman S. and Geman D. 1984. Stochastic relaxation, Gibbs distribution, and the Bayesian restoration of images. IEEE Transactions on Pattern Analysis and Machine Intelligence 6: 721–741.

    Article  MATH  Google Scholar 

  • Geyer C.J. and Thompson E.A. 1992. Constrained Monte Carlo maximum likelihood for dependent data (with discussion). Journal of Royal Statistical Society, Series B 54: 657–699.

    MathSciNet  Google Scholar 

  • Gu M.G. and Kong F.H. 1998. A stochastic approximation algorithm with Markov chain Monte Carlo method for incomplete data estimation problems. In: Proceeding of National Academic Science of USA 95: 7270–7274.

    Article  MATH  MathSciNet  Google Scholar 

  • Gu M.G. and Zhu H.T. 2001. Maximum likelihood estimation for spatial models by Markov chain Monte Carlo stochastic approximation. Journal of Royal Statistical Society, Series B 63: 339–355.

    Article  MATH  MathSciNet  Google Scholar 

  • Higdon D.M. 1998. Auxiliary variable methods for Markov chain Monte Carlo with applications. Journal of American Statistical Association 93: 585–595.

    Article  MATH  Google Scholar 

  • Horton N.J. and Laird N.M. 1998. Maximum likelihood analysis of generalized linear models with missing covariates. Statistical Methods in Medical Research 8: 37–50.

    Article  Google Scholar 

  • Huang F. and Ogata Y. 2001. Comparison of two methods for calculating the partition functions of various spatial statistical models. The Australian and New Zealand Journal of Statistics 43: 47–65.

    Article  MATH  MathSciNet  Google Scholar 

  • Huffer F.W. and Wu H.L. 1998. Markov chain Monte Carlo for auto-logistic regression models with application to the distribution of plant species. Biometrics 54: 509–524.

    Article  MATH  Google Scholar 

  • Jens L.J. and Niels V.P. 1999. Asymptotic normality of the maximum likelihood estimator in state space models. Annals of Statistics 27: 514–535.

    Article  MATH  MathSciNet  Google Scholar 

  • Karcher P. and Wang Y. 2001. Generalized nonparametric mixed effects models. Journal of Computational and Graphical Statistics 10: 641–655.

    Article  MathSciNet  Google Scholar 

  • Kwan R.K.S., Evans A.C., and Pike G.B. 1999. MRI simulation-based evaluation of image-processing and classification methods. IEEE Transactions on Medical Imaging 18: 1085–1097.

    Article  Google Scholar 

  • Lai T.L. 2003. Stochastic approximation. Annals of Statistics 31: 391–406.

    Article  MATH  MathSciNet  Google Scholar 

  • Lange K. 1995. A gradient algorithm locally equivalent to the EM algorithm. Journal of Royal Statistical Society, Series B 55: 425–437.

    Google Scholar 

  • Lee Y. and Nelder J.A. 1996. Hierarchical generalized linear models (with discussion). Journal of Royal Statistical Society, Series B 58: 619–678.

    MATH  MathSciNet  Google Scholar 

  • Li S.Z. 2001. Markov Random Field Modeling in Image Analysis. Springer-Verlag, Tokyo.

    MATH  Google Scholar 

  • Liu J. 2001. Monte Carlo Strategies in Scientific Computing. Springer, New York.

    MATH  Google Scholar 

  • Louis T.A. 1982. Finding the observed information matrix when using the EM algorithm. Journal of Royal Statistical Society, Series B 44: 190–200.

    MATH  MathSciNet  Google Scholar 

  • Marroquin J.L., Santana E.A., and Botello S. 2003. Hidden Markov measure field models for image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence 25: 1380–1397.

    Article  Google Scholar 

  • McCullagh P. and Nelder J.A. 1989. Generalized Linear Models (2nd edn.). Chapman and Hall, London.

    MATH  Google Scholar 

  • Møller J. 1999. Markov chain Monte Carlo and spatial point processes. In W.S. Kendall, O.E. Barndorff-Nielsen, and M.C. van Lieshout (Eds.), Stochastic Geometry: Likelihood and Computation, Chapman and Hall, London.

  • Moyeed R.A. and Baddeley A.J. 1991. Stochastic approximation of the maximum likelihood estimate for a spatial point pattern. Scandinavian Journal of Statistics 18: 39–50.

    MATH  MathSciNet  Google Scholar 

  • Ortega J.M. 1990. Numerical Analysis: A Second Course. Society for Industrial and Academic Press, Philadelphia.

    MATH  Google Scholar 

  • Penttinen A. 1984. Modelling interaction in spatial point patterns: parameter estimation by the maximum likelihood method. Jy. Stud. Comput. Sci. Econometr. Statist. 7.

  • Pettitt A.N., Friel N., and Reeves R. 2003. Efficient calculation of the normalisation constant of the autologistic model on the lattice. Journal of Royal Statistical Society, Series B 65: 235–247.

    Article  MATH  MathSciNet  Google Scholar 

  • Polyak B.T. 1990. New stochastic approximation type procedures. Autom. Telem. pp. 98–107. (English translation in Automat. Remote Contr. 51).

  • Polyak B.T. and Juditski A.B. 1992. Acceleration of stochastic approximation by averaging. SIAM Journal of Control and Optimization 30: 838–855.

    Google Scholar 

  • Qian W. and Titterington D.M. 1991. Estimation of parameters in hidden Markov models. Philosophical Transactions of the Royal Society of London, Series A 337: 407–428.

    MATH  Google Scholar 

  • Rajapakse J.C., Giedd J.N., and Rapoport J.L. 1997. Statistical approach to segmentation of single-channel cerebral MR images. IEEE Transactions on Medical Imaging 16: 176–186.

    Article  Google Scholar 

  • Robbins H. and Monro S. 1951. A stochastic approximation method. Annals of Mathematical Statistics 22: 400–407.

    MathSciNet  MATH  Google Scholar 

  • Robert C.P. and Casella G. 1999. Monte Carlo Statistical Methods. Springer-Verlag, New York.

    MATH  Google Scholar 

  • Rydén T. 1997. On recursive estimation for hidden Markov models. Stochastic Processes and their Applications 66: 79–96.

    Article  MATH  MathSciNet  Google Scholar 

  • Saquib S.S., Bouman C.A., and Sauer K. 1998. ML parameter estimation for Markov random fields with applications to Bayesian tomography. Transactions on Image Processing 7: 1029–1044.

    Article  Google Scholar 

  • Stoer J. and Bulisch R. 1980. Introduction to Numerical Analysis. Springer-Verlag, New York.

    Google Scholar 

  • Swendsen R.H. and Wang J.S. 1987. Nonuniversal critical dynamics in Monte Carlo simulation. Physics Review Letters 58: 86–88.

    Article  Google Scholar 

  • Wang N., Lin X., Gutierrez R.G., and Carroll R.J. 1998. Bias analysis and SIMEX approach in generalized linear mixed measurement error models. Journal of American Statistical Association 93: 249–261.

    Article  MATH  MathSciNet  Google Scholar 

  • Wei G.C.G. and Tanner M.A. 1990. A Monte Carlo implementation of the EM algorithm and the Poor man’s data augmentation algorithm. Journal of American Statistical Association 85: 699–704.

    Article  Google Scholar 

  • Winkler G. 1995. Image Analysis, Random Fields and Dynamic Monte Carlo Methods: A Mathematical Introduction. Springer-Verlag, Berlin Heidelberg.

    MATH  Google Scholar 

  • Younes L. 1989. Parameter estimation for imperfectly observed Gibbsian fields. Probability Theory Related Feilds 82: 625–645.

    Article  MATH  MathSciNet  Google Scholar 

  • Zeger S.L. 1988. A regression model for time series of counts. Biometrika 75: 621–629.

    Article  MATH  MathSciNet  Google Scholar 

  • Zeger S.L., Liang K.Y., and Albert P.S. 1988. Models for longitudinal data: a generalized estimating equation approach. Biometrics 44: 1049–1060.

    Article  MATH  MathSciNet  Google Scholar 

  • Zhang H. 2002. On estimation and prediction for spatial generalized linear mixed models. Biometrics 56: 129–136.

    Article  Google Scholar 

  • Zhang Y., Brady M., and Smith S. 2001. Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Transactions on Medical Imaging 15: 45–57.

    Article  Google Scholar 

  • Zhu H.T. and Gu M.G. 2005. Maximum likelihood from spatial random effects models via the stochastic approximation expectation maximization algorithm (supplement). Technical report, Department of Biostatistics, University of North Carolina at Chapel Hill.

  • Zhu H.T. and Lee S.Y. 2002. Analysis of generalized linear mixed models via a stochastic approximation algorithm with Markov chain Monte Carlo method. Statistics and Computing 12: 175–183.

    Article  MathSciNet  Google Scholar 

  • Zhu H.T. and Zhang H.P. 2004. Hypothesis testing in a class of mixture regression models. Journal of Royal Statistical Society, Series B: 66: 3–16.

    Article  MATH  MathSciNet  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Biostatistics and Biomedical Research Imaging Center, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7420, USA

    Hongtu Zhu

  2. Department of Psychiatry, Columbia University and New York State Psychiatric Institute, 1051 Riverside Drive, Unit 74, New York, New York, 10032, USA

    Minggao Gu

  3. Department of Statistics, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, P.R. China

    Bradley Peterson

Authors
  1. Hongtu Zhu
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Minggao Gu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Bradley Peterson
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Hongtu Zhu.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Zhu, H., Gu, M. & Peterson, B. Maximum likelihood from spatial random effects models via the stochastic approximation expectation maximization algorithm. Stat Comput 17, 163–177 (2007). https://doi.org/10.1007/s11222-006-9012-9

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11222-006-9012-9

Keywords

Search

Navigation